# MartialBE/one-hub

OpenAI 接口管理 & 分发系统，改自songquanpeng/one-api。支持更多模型，加入统计页面，完善非openai模型的函数调用。

Repository: https://github.com/MartialBE/one-hub
Canonical: https://ross.abutalabs.com/products/one-hub
Homepage: https://one-hub.xiao5.info/
Language: Go
License: Apache-2.0
License Family: permissive
Topics: chatgpt, openai, openai-api
Last push: 2026-02-19T21:16:31+00:00

## Health v2 (maintenance only)
Score: 67/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 68, release rhythm 60, longevity 77
- inputs: {"age_days": 1089, "days_push": 195, "days_rel": 270, "gap_med": 10.0, "n_releases_24m": 35}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2882, forks 495 (observed 2026-08-28T04:07:28.042274+00:00)

## What it is
One Hub is a self-hosted OpenAI-compatible API management and distribution system forked from songquanpeng/one-api, written in Go. It aggregates multiple LLM providers (OpenAI, Claude, Gemini, and more) behind a unified API with user management, billing, statistics dashboards, and improved function-calling support for non-OpenAI models.

## Use cases
- manage and distribute openai api keys across a team
- unify multiple llm providers behind one openai-compatible endpoint
- self-host an llm api gateway with usage billing
- proxy claude and gemini requests through an openai-style api
- track token usage and costs per user with dashboards
- add function calling support to non-openai models
- resell llm api access with quotas and user groups

## When to choose
- you need a single OpenAI-compatible gateway in front of many LLM providers
- you want per-user quotas, billing, and usage analytics for LLM access
- you need better function-calling compatibility for non-OpenAI models than one-api offers
- you want Prometheus monitoring and channel-level proxying for LLM traffic

## When to avoid
- you need guaranteed stability or official support - the project is maintained for personal learning with no support guarantees
- you want to migrate an existing one-api deployment - the database schema is incompatible
- you only use a single LLM provider directly without needing distribution or key management
- you must serve unregistered generative AI services to the public in China, which the license terms prohibit

## Facets
- artifact type: service
- maturity: active
- function: api-gateway, llm-inference, proxy, monitoring, auth
- domain: large-language-models, apis, self-hosted, developer-tools, artificial-intelligence
- platform: self-hosted, go, cross-platform
- tags: openai-api, api-distribution, llm-gateway, one-api-fork, token-management, multi-provider, rate-limits, telegram-bot, prometheus, billing, docker, web-server, linux

## Member repositories
- MartialBE/one-hub (main) score 67

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:28.042274+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T18:48:25.526706+00:00, confidence not recorded.
  - readme: https://github.com/MartialBE/one-hub (fetched 2026-08-28T04:07:28.042274+00:00, sha 4959d5be683f)
  - homepage: https://one-hub.xiao5.info/ (fetched 2026-08-29T09:50:52.398732+00:00, sha c32135ae723c)
- Data as of 2026-08-30T08:39:29.467469+00:00.